Fertility Awareness Methods

Dynamic model for multivariate markers of fecundability

Cai B, Dunson DB, Stanford JB

Published September 2010 Biometrics, 66(3), 905-913
DOI 10.1111/j.1541-0420.2009.01327.x PMID 19751248

RRM Academy Synopsis

A model tracks cervical mucus stretch and color across the cycle

Researchers built a Bayesian model that follows cervical mucus stretch and color through the whole menstrual cycle, using daily Creighton Model records from 323 women. The model estimated each cycle's peak mucus day and ovulation day, with wide variation between women and cycles. The authors report good fit.

Key Findings

  • The data covered 1676 cycles from 323 women, aged 18 to 39 years. Those cycles included 115 conceptions.
  • The rounded posterior mean for ovulation day was the 19th day (95% credible interval 14, 24) by mucus stretch and the 18th day (14, 23) by mucus color.
  • For both characteristics, the most likely peak mucus day came a bit before ovulation, and the stretch peak fell slightly after the color peak.
  • Goodness-of-fit values were 0.54 and 0.057, which the authors read as good fit of the proposed model.

Interpretation

The paper is a statistical methods report. The authors applied the model to a retrospective cohort of women who kept Creighton Model FertilityCare System (CrMS) charts, with daily standardized mucus observations. The model treats each woman's and each cycle's peak mucus day and ovulation day as latent quantities, so the fertile and infertile windows carry uncertainty across women and cycles. It estimates the trajectory of mucus stretch and color across the entire cycle and borrows strength on the average trajectory when observations are missing. The authors name one limitation: the model fixes the length of the fertile interval.

RRM Context

Cervical mucus observations anchor Creighton Model charting. The paper presents a model of how those observations shift over whole cycles and of the spread in peak mucus day and ovulation day among women. Its later peak for stretch than for color agrees with earlier clinical reports that color often shifts before stretch does as mucus turns less fertile.

Abstract

Dynamic latent class models provide a flexible framework for studying biologic processes that evolve over time. Motivated by studies of markers of the fertile days of the menstrual cycle, we propose a discrete-time dynamic latent class framework, allowing change points to depend on time, fixed predictors, and random effects. Observed data consist of multivariate categorical indicators, which change dynamically in a flexible manner according to latent class status. Given the flexibility of the framework, which incorporates semi-parametric components using mixtures of betas, identifiability constraints are needed to define the latent classes. Such constraints are most appropriately based on the known biology of the process. The Bayesian method is developed particularly for analyzing mucus symptom data from a study of women using natural family planning.

Read the guide

  • The Billings Ovulation Method (Billings Method)

    The Billings Ovulation Method uses only cervical mucus to identify the fertile window. Learn how it works, what the evidence shows, and its value in restorative reproductive medicine.

    12 min read

Topics

By this author

Related research

Fertility Awareness › Methods › Billings Ovulation Method · Menstrual Cycle › Cycle Biomarkers › Cervical Mucus · Research Methods › Measurement and Statistics › Statistical Methods
Bo Cai, David B Dunson, Joseph B Stanford
B Cai, Dave Dunson, D Dunson, Joe Stanford, Joey Stanford, J Stanford
PMID 19751248 19751248 DOI 10.1111/j.1541-0420.2009.01327.x 10.1111/j.1541-0420.2009.01327.x Cai et al. 2010, Cai 2010